Author: Melvin S. Schwartz
Date: 17:20:56 06/30/99
Go up one level in this thread
On June 30, 1999 at 19:40:50, Steve Lopez wrote: >On June 30, 1999 at 19:15:00, Melvin S. Schwartz wrote: > > >>Would you please explain to me why one should have to "reset" the tree? > >Hiarcs and Fritz re-weight their opening preferences based on their prior >success or failure in various openings. For example, if you play Hiarcs (or >Fritz) 10 games in a certain branch of, say, the Sicilian, and it beats you all >ten games, it will probably tend to prefer that branch over all others. This >branch might be a *great* line to play against you (posibly due to your lack of >familiarity with that opening), but it might be a lousy opening to play against >another computer program or a very strong human opponent. By resetting the >weights, you restore the tree to its default setting (making Hiarcs or Fritz >"forget" its previous results in that line). See section 4.6 "Book options" and >section 4.7 "Reset weights" in the Fritz5 manual. > >>After >>copying the Hiarcs opening tree to the hard drive, why would it be necessary to >>repeat the process? > >You don't have to repeat the process. "Reset weights" has absolutely nothing to >do with recopying the book to the hard drive. > >-- Steve Lopez ________________________________________________________ I presented two openings Hiarcs played that I do believe were awful. I think this is what led to the issue of the quality in Hiarcs opening book. Terry Ripple, I believe, said that his Hiarcs 7.32 wouldn't play the Sicilian line I presented and that it must be because tournament book was off or the tree had to be reset. I believe that is the story here. Now, my Hiarcs 7.32 played the line at 40/2 with tournament book checked on and I cannot understand why his Hiarcs wouldn't play it unless he modified the tree, which he did not say was done. Therefore, something is amiss here. The line for you to peruse is: Hiarcs: Black: 1.e4...c5 2.Nf3...e6 3.d4...cxd4 4.Nxd4...Nf6 5.Nxc6...bxc6 6.e5...Nd5 (he claims his Hiarcs would not play Nd5). 7.Ne4...f5 8.Nd6+...and here while the screen info said Bxd6 no, with no highlighted in red, Hiarcs played Bxd6. However, that wasn't the issue - it was that Nd5. In any event, after the 7 moves, I believe Hiarcs has put itself in a fairly bad position by not playing the proper book line of Qc7 instead of Nd5. I also have another opening example with Hiarcs as White playing an awful first 8 moves where Black gets connected passed pawns in the center. This is opening book I'm referring too. Regards, Mel
This page took 0.01 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.